使用Miniconda3运行Stable Diffusion时遇导入错误求助

问题描述 #

按照线上教程部署本地AI图像生成器，执行图像生成指令时出现如下报错：

File "scripts/txt2img.py", line 22, in <module>
    from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
  File "D:\miniconda3\envs\ldm\lib\site-packages\diffusers\__init__.py", line 29, in <module>
    from .pipelines import OnnxRuntimeModel
  File "D:\miniconda3\envs\ldm\lib\site-packages\diffusers\pipelines\__init__.py", line 19, in <module>
    from .dance_diffusion import DanceDiffusionPipeline
  File "D:\miniconda3\envs\ldm\lib\site-packages\diffusers\pipelines\dance_diffusion\__init__.py", line 1, in <module>
    from .pipeline_dance_diffusion import DanceDiffusionPipeline
  File "D:\miniconda3\envs\ldm\lib\site-packages\diffusers\pipelines\dance_diffusion\pipeline_dance_diffusion.py", line 21, in <module>
    from ..pipeline_utils import AudioPipelineOutput, DiffusionPipeline
  File "D:\miniconda3\envs\ldm\lib\site-packages\diffusers\pipelines\pipeline_utils.py", line 67, in <module>
    from transformers.utils import SAFE_WEIGHTS_NAME as TRANSFORMERS_SAFE_WEIGHTS_NAME
ImportError: cannot import name 'SAFE_WEIGHTS_NAME' from 'transformers.utils' (D:\miniconda3\envs\ldm\lib\site-packages\transformers\utils\__init__.py)

问题原因 #

不是操作步骤出错，而是diffusers和transformers两个Python库的版本不兼容：SAFE_WEIGHTS_NAME是transformers较新版本才引入的常量，当前安装的diffusers版本需要这个常量，但环境里的transformers版本过低，导致无法导入。

解决方法 #

• 方法1：直接升级transformers
  激活conda环境（ldm）后执行命令：

  pip install --upgrade transformers
  

• 方法2：指定兼容版本对
  如果升级后仍有问题，可安装经过验证的兼容版本组合：

  pip install diffusers==0.20.2 transformers==4.31.0
  

  若用conda管理环境，也可执行：

  conda install -c conda-forge transformers==4.31.0 diffusers==0.20.2
  

• 方法3：检查环境依赖
  执行以下命令查看当前环境的依赖版本，确认是否存在其他冲突：

  pip list | grep -E "diffusers|transformers"
  

内容的提问来源于stack exchange，提问作者JB24